I understand what you are saying.... I don't think we disagree. There is no longer a need for mobile to have its own identity.
Technically, yes, in every other way (economically, branding, etc.), I think not, but I'll skip that for now.
My impression is that you may be confusing the 'means' with the 'ends'. The author is definitely doing so.
Nothing related to Win32 will ever get a standing ovation. Win32 (what most people actually associate with Windows) is something MOST people MUST use rather than something they WANT to use.
All desktop OSes are that way, as they were developed at a time when computers were isolated/unconnected boxes and their users were expected to understand technology. That is no longer the case, which is why mobile OSes have become the so much more successful mass market products. Mobile OSes (or consumerized OSes) were explicitly designed to require zero administration/maintenance and to protect users from themselves. Win32 can't and will never be able to do that (without sacrificing compatibility), which is one of the reasons MS broke from it and designed what amounts to a second platform, WinRT, which evolved into UWP.
For the reasons mentioned, MS needs UWP to be successful if they want to remain relevant as a platform provider. Windows Phone was intended to be the vehicle to popularizing the UWP, but that failed. Now, Win32 is the only thing MS has which could carry UWP into the mass market, as MS apparently can't design UWP in a way were it can achieve success on its own merits (MS' biggest problem).
So, we're not getting a small device that runs Win32 software because that is what MS always wanted. We could have easily had that 6 years ago. We're getting it because it's the only half-viable path MS still has to popularizing the UWP. It's a move out of weakness/necessity, rather than strength/innovation.
The goal is to have UWP on everything and Win32 in the dustbin. Win32 on small devices is a means to an end (make UWP a viable and even the preferred platform for software development on all form factors), not the end in and of itself.